An Infrared Study of the Metal Chelates of Some Imidazole

An Infrared Study of the Metal Chelates of Some Imidazole Derivatives1. Thomas R. Harkins, Joseph L. Walter, Otho E. Harris, and Henry Freiser. J. Am...
0 downloads 0 Views 627KB Size
260

T. R.HARKINS, J. L. WALTER,OTHO E. HARRIS [CONTRIBUTIONFROM

THE

AND

HENRYFREISER

Vol. 78

DEPARTMENT OF CHEMISTRY, UNIVERSITY O F PITTSBURGH]

An Infrared Study of the Metal Chelates of Some Imidazole Derivatives1 BY THOMAS R. HARKINS, JOSEPH L. WALTER, OTHOE. HARRISAND HENRYFREISER RECEIVED JUNE 16, 1955 Infrared absorption curves have been obtained in the 2-15 p region for a series of metal chelates of 2-(2-pyridyl)- and 2(o-hydroxypheny1)-derivativesof benzimidazole, imidazoline, benzoxazole and benzothiazole. Use was made of the potassium bromide pressed pellet technique for the preparation of samples. Particular attention has been focussed on the 3 j~ region in order to observe the effects of coordination on the N-H stretching frequency. Coordination of the reagent with a metal ion decreases the N-H stretching frequency. This is explained in terms of resonance structures. Evidence is given for an intramolecular chelate structure in 2-(o-hydroxypheny1)-benzimidazole.2-( o-Hydroxypheny1)-benzoxazole and its chelates have an absorption band a t 8.0 p (1250 kayser) which undergoes a regular shift, paralleling the usual stability order of the metals. The copper(I1) chelate is a t the highest frequency. 2-(2-Pyridyl)-imidazoline was prepared by the procedure outlined in a paper by Walter and Freiser.8 The procedure for the preparation of 2-(2-pyridyl)-benzoxazole was analogous to that followed in the preparation of 2-(2-pyridyl)-benzimidazole. One mole of or-picolinic acid and one mole of o-aminophenol were heated together for three hours in a distillin: flask immersed in an oil-bath a t a temperature of 160-170 . The product was then distilled a t atmospheric pressure. The product was recrystallized three times from an alcoholwater mixture to giv: the 2-(2-pyridyl)-benzimidazolewhich N, melted a t 104-106 . Anal. Calcd. for CIZH~NZO: 14.25. Found: N, 14.40. The preparation of 2-(o-hydroxyphenyl)-imidazoline is given in a thesis by W. D. Johnston.9 Equimolar amounts of salicylamide and ethylenediamine were heated together in a distilling flask a t 140-150 a for tw: \3 // hours after which the temperature was raised to 175-185 S and the heating continued for another two hours. Both and ammonia were evolved. The mixture was then where X-X is either CHz-CH2 or CeH4; Y is either water distilled a t approximately 330 '. After decolorization with N-H, 0, or S; and 2 is either N or C-OH. charcoal and recrystallization from an alcohol-water mixI n those compounds where 2 was C-OH, che- ture, the product melted a t 200-203 Solutions of each of the aforementioned organic reagents lation was accompanied by a neutralization of having a concentration of 20 mg. per ml. were prepared in charge while for those in which 2 was N, chelation 95y0 ethanol and used in aliquots as needed to form the varioccurred between the two pyridyl nitrogen atoms ous metailic complexes. Reagent grade nitrates of copper(II), nickel(II), cobaltand the metal with no charge neutralization. The potassium bromide pressed disk technique2s3 ( I I ) , zinc(II), cadmium(II), lead(II), manganese(II), magnesium(I1) and iron(I1) were used to prepare solutions which has been used to obtain the spectra of these solid contained 50 mg. of metal per ml. for use in preparing the metal complexes. various metallic chelates. Considerable interest centered around the 3 p Ten per cent. solutions of reagent grade sodium perregion of the spectra (N-H stretching vibration re- chlorate, potassium bromide and potassium iodide were used to precipitate the metallic complexes of 2-(2-pyridyl)gion) of those compounds containing an N-H group benzimidazole and 2-(2-pyridyl)-imidazoline. since a potentiometric study of some substituted Potassium bromide powder for making disks was prepared imidazole complexes4 had produced evidence of the as follows. Reagent grade crystalline potassium bromide was ground dependence of the acidity of the N-H grouping in a motor-driven mortar for several hours. The resultant upon the metal ion present. Depending on the powder was dried a t 130" for three days and then kept in a nature of the nitrogen-to-metal linkage, one might desiccator until used. expect some type of electronic interaction through Apparatus.-A one inch (i.d.) metallurgical mold assembly the imidazole ring system that will in some way obtained from Buehler Co., Chicago, Illinois, was used in conjunction with a Thiele testing apparatus with a capacity affect the strength of the N-H bond. of 60,000 p.s.i. to press the potassium bromide disks. A Model 12-C Perkin-Elmer Recording Infrared SpecExperimental trometer equipped with sodium chloride optics was used for Materials .-2-( o-Hydroxyphenyl)-benzoxazole was pre- all infrared measurements. The frequencies given are accurate to a t least kt10 kayser (f0.01p ) in the 3 p region. pared by the method outlined by Siegfried and M ~ s e r . ~ Preparation of Metal Chelates.-A number of qualitative 2-(o-Hydroxyphenyl)-benzimidazole was prepared by the observations were made using the various reagents and the procedure outlined in a paper by Walter and Freiser.O 2-(2-Pyridyl)-benzimidazole was prepared according to metal ions mentioned under materials. The pH was varied to obtain the optimum conditions for precipitation. Also, the procedure given by Lek0 and Llajinats.' in the cases of the reagents forming charged, soluble com(1) Presented in part a t the Pittsburgh Conference on Analytical plexes with the metals, several anions, e.g., perchlorate, Chemistry and Applied Spectroscopy, March, 1955. chloride, iodide and bromide, were tried in an effort to pre(2) M. M. Stimson and h l . J . O'Donnell, THISJOURNAL, 74, 1805 cipitate the compounds so as to obtain solid chelates for the (1952). infrared investigation. (3) U.Schiedt and H. Reinwein, 2. Katrrufororsch.,Tb, 270 (1950). The following procedures represent the optimum condi(4) T. R. Harkins and H. Freiser, THISJOURNAL, in press. tions found for the preparation of the chelates used in this (5) S. Siegfried and M. Moser, Ber., 66, 1089 (1922). study.

Introduction This paper deals with an infrared spectroscopic study of a series of related heterocyclic compounds and their metal chelates which was undertaken to investigate the effects of metal bonding upon the characteristic absorption frequencies of the compounds. The compounds included in this work may all be represented by the skeletal structure

".

(6) J. L. Walter and H. Freiser, Anal C h e n . , 26, 127 (1953). (7) A. Lek0 and G. Vlajinats Bull. S O L . chim. Y O Y . Bougoslau., 1, 3 (1930).

( 8 ) J . L. Walter and H. Freiser, Anal. Ckcm., a6, 217 (1954). (9) W. D.Johnston, Ph.D. Thesis, University of Pittsburgh, 1953.

Jan. 20, 1956

STUDY OF METALCHELATES OF IMIDAZOLE DERIVATIVES

Chelates of 2-(o-Hydroxypheny1)-benzoxazole.-One milliliter of each of the metal ions to be precipitated as a chelate was pipetted into small beakers. Ten milliliters of distilled water was then added to each. An aliquot of the reagent solution containing the amount necessary to precipitate each of the metals was then added to each of the samples. The pH of each solution was adjusted to the optimum pH as determined from the qualitative reactions. For Cu the optimum PH was 4,for Cd 10, and for Ni,Co, etc., a value of 7 . The precipitates were filtered and washed several times with 50% ethanol-water solution. The precipitates were then transferred to small sample vials and dried a t 130" for several hours. This was the only purification procedure given these chelates. Since they were so insoluble, no solvent could be found to offer any suitable means of recrystallization. Chelates of 2-(o-Hydroxyphenyl)-benzimidazole.-Essentially the same procedure as outlined above for the preparation of the chelates of 2-(o-hydroxyphenyl)-benzoxazole was used to obtain the chelates of 2-(o-hydroxyphenyl)-benzimidazole. However, from the qualitative tests made with this reagent, it was discovered that only the chelates of copper(II), cobalt(I1) and zinc(I1) could be prepared readily. The copper(I1) chelate was precipitated in a neutral solution, while the cobalt(I1) and zinc(I1) chelates were precipitated in a slightly basic solution (pH 8). Chelates of 2-(o-Hydroxyphenyl)-imidazoline.-The chelates of this reagent were also prepared in the same manner as the above-mentioned chelates. The qualitative tests indicated that it was possible to obtain as solids only the copper(11), nickel(II), cobalt(I1) and zinc(I1) chelates. The copper( 11) chelate was precipitated from a neutral solution, and the cobalt(II), nickel(I1) and zinc(I1) chelates were precipitated from a slightly basic solution. Chelates of 2-(2-Pyridyl)-benzimidazole.-From qualitative tests with metals and this reagent it was found that all of the metals formed soluble, charged complexes with the reagent much in the same manner as 2,2'-bipyridine and o-phenanthroline. Thus, a suitable anion that would give a precipitate had to be sought. From the group of anions tried, it was found that the perchlorate, bromide and iodide ions gave precipitates with copper(II), cobalt(II), nickel(I1) and zinc(I1) complexes. One milliliter each of the metal ion solutions of copper(II), nickel(II), cobalt(I1) and zinc(I1) was pipetted into small beakers. Ten milliliters of distilled water was added. An aliquot of the reagent solution containing the amount necessary to complex the metal was added. The pH was adjusted to about six. Then about five milliliters of the prepared sodium perchlorate solution was added to each sample. The perchlorates of the complexes immediately separated from solution. The precipitates were filtered and recrystallized from hot water. The recrystallized complex was filtered and washed several times with cold 30% alcoholwater solution. These precipitates were dried a t 130'. In order to prepare the iodides and bromides of the metal complexes of 2-(2-pyridyl)-benzimidazole,exactly the same procedure as was outlined for the precipitation of the perchlorates was utilized except that the precipitating anion was either the prepared iodide or bromide solution in place of the perchlorate solution. Chelates of 2-(2-Pyridyl)-imidazoline.-Exactly the same procedure used to prepare the perchlorates of the metal complexes of 2-(2-pyridyl)-benzimidazole was employed to prepare the chelates of 2-(2?-pyridyl)-imidazoline. For this study the 2-(2-pyridyl)-imidazolineiodides of copper(11), nickel(II), cobalt(I1) and zinc(I1) were prepared as well as the imidazoline bromides of the same metals. The imidazoline complexes are much more soluble in an alcoholic-water solution than are the 2-(2-pyridyl)-benzimidazole complexes, and so care must be exercised in the washing of the chelates. The precipitates are dried a t 130". The results of microanalyses are given in Table I for those compounds having the expected stoichiometry. Preparation of Potassium Bromide Sample Disks.Since the chelates as a whole were insoluble in organic solvents, it was necessary to employ a method for handling solids. The potassium bromide pellet technique was This technique rather than the oil mull technique chosen. was used as paraffin oil has a strong absorption band in the 3 p region, which would be undesirable in this study. For the qualitative type work undertaken in this investigation, it was found that the following method for the *t3

26 1

preparation of the potassium bromide sample disks was most adequate. A sample of from 5-7 mg. of the compound to be studied, along with 800 mg. of the prepared potassium bromide powder, was thoroughly mixed and ground together in an agate mortar. This grinding took about three minutes per sample. The thoroughly mixed sample-potassium bromide preparation was then transferred into the mold and distributed evenly. A pressure of 24,000 p.s.i. was applied t o the mold for ten minutes. Although entirely transparent disks were not produced by this method, disks with a sufficiently large enough transparent area for study were obtained. In so far as this study could reveal, the prepared discs can be stored in a desiccator for an indefinite period of time before an infrared curve is obtained for the sample. Several curves run over a period of several weeks on the same disk showed no marked differences in any of the absorption peaks.

Discussion 3 p Absorption Bands (Table I).-2-(2-Pyridyl)benzimidazole has a broad absorption band in the 3 p region (3040 kayser), which is doubtless due to the N-H stretching vibration. A spectrum of 2-(2pyridyl) -benzoxazole, an identical molecule except that an oxygen atom replaces the N-H group, does not show an absorption band in this region. The broadness and tapering off of the absorption band in the direction of lower frequencies is indicative of a strong hydrogen bonding effect operating on the N-H bond. It is not surprising to find this effect since benzimidazoles and imidazoles possessing a free imino hydrogen are known to be capable of associating. lo The association is predominantly intermolecular in nature rather than intramolecular since a dilute solution of the compound in chloroform has a very sharp peak a t a much higher frequency (3435 kayser) . Sutherland" has suggested that hydrogen bonding should be considered only for bands in which A:/e > 3% since a change in state will also result in a change in frequency. From the data of Table I one can calculate band differences of 12 and 6% for 2-(2-pyridyl) -benzimidazole and 2-(2-pyridyl)imidazoline, respectively. Depending on the packing of the molecules in the solid state, any one or all of the three nitrogens in the molecule could enter into hydrogen bonding. The solid metal chelates of 2-(2-pyridyl)-benzimidazole have their N-H absorption peaks around 3000 kayser. Either there still is considerable intermolecular hydrogen bonding in the solid because of the close packing of the molecules, or else because of chelation there is a pronounced shift of the free N-H absorption to lower frequencies. Since the two basic pyridine nitrogens of the reagent are being utilized for complex formation, one would expect a lesser degree of hydrogen bonding t o occur between molecules of the chelates than between molecules of the reagent itself. I n no event is the N-H absorption frequency of the chelate a t a higher frequency than the band observed for the solid reagent. One might possibly infer then that coordination with a metal results in a decrease of the N-H stretching frequency. The same effect is noted for the chelates of 2-(2pyridyl) -imidazoline. (10) K . Hofmann, "Imidazole and Its Derivatives," Interscience Publishers, Inc., New York, N . Y.,1953. (11) G. B. B. M . Sutherland, Trans. Faraday SOC. 36,889 (1940).

T. R.HARKINS, J. L. WALTER, OTHOE. HARRISAND HENRYFREISER

262

3p

Vol. 7s

TABLE I ABSORPTION FREQUENCIES AND ANALYSES Stretching frequencies (kayser) e N-H CHz 0-H

Formula

Compound

CizHgNa

3040 3435" Ni( C1zHeNs)aIz 3030 CO(CizH9Na)sI:: 3010 Zn(CizHeNa)zIz 3005 C U ( C ~ Z H S W Z B ~ Z 3040 Ni( C I Z H S N S ) ~ B ~ Z 2995 CO(CIZHSN&B~Z 3005 Ni( C12H9N3)2.(C10& 3005 C O ( C ~ Z H O N ~ ) ~ ( C3020 ~O~Z CsH9N3 3215 3415b

2-(2-Pyridyl)-benzimidazole Chelates

2-(2-Pyridyl)-imidazoline

Chelates

Ni( CaHgNdJ, Co(CaHsNs)aIz Zn( CsHeN8)zBr~

2-( o-Hydroxypheny1)-benzimidazole Chelates

3175 3175 3085 3210 3155 2985 3105 2995 2995 3065 3080 3155 3080

ClsHloNzO CU(Ci3HgNzO)z CO(Ci3HgNzO i z Z~(CI~H&O)Z 2(-(o-Hydroxyphenyl)-imidazoline CgHd?ZO Chelates C~(c9HoNzO)z Xi( COHgN20)2 Co(CeHoN~0)z Zn(CgH9NzO)z 2 4 o-Hydroxyphenyl)-benzoxazole Ci3H9NOz .. Chelates CU(Ci3HsNOz)z Ni( C I ~ H ~ N O Z ) Z CO(C~~HSNO& Zn( C 1 3 H W " Pb( Ci3HaNOz)z Mg( CisH~N0zh Cd( CiaHsN0n)z Mn(C13HsN0zh Kayser = ern.-'. Dilute solution in CCI,. Dilute solution in CHCla.

Considering now the reagents having an o-hydroxyphenyl-group as the substituent in the 2-position, it was found that 2-(o-hydroxyphenyl)benzimidazole has a broad absorption band centered a t about 2530 kayser, which would probably be due to the 0-H stretching vibration. The corresponding 2-(2-pyridyl)-benzimidazole,which has no 0-H bond, does not show this absorption. Since the 0-H absorption band of this reagent is a t such a low frequency, this would be evidence for the existence of a "chelated" 0-H bond in the reagent itself. Hunter and Marriott12 have postulated this structure from a consideration of its lack of molecular association. The differences in the ultraviolet absorption spectra of the 0 - and m- and p-hydroxyphenyl derivatives of benzimidazole have also been shown.13 H

I

0:>-a N

I

hT.. . H-0

(12) L. Hunter and J. A. Marriott, J . Chem. Soc., 777 (1941). (13) Chr. Wiegand and E.Merkef, A n n . , 167, 242 (1947).

..

Nitrogen, 7, Calcd. Found

..

13.69

13.85

2900 2855b 2930b 2900 2900 2900

..

14.04 14.03 11.84 13.69 15.68 15.67 14.95 14.95 28.55

13.78 13.85 12.18 13.85 15.45 15.52 14.51 14.71 28.32

16.72 18.71 16.18

16.41 16.72 16.27

..

2530

..

..

2845 2845 2845 2845 2845

2655

..

..

13.32 11.63 11.74 11.58 17.28 14.52 14.71 14.70 14.45 5.25 5.79 5.85 5.85 5.77 4.46 6.30 5.26 5.89

13.62 11.34 12.17 12.01 17.23 14.79 15.03 14.82 13.81 5.21 6.24 6.37 6.26 6.14 4.93 6.60 5.21 6.28

..

There is a distinct difference in the 3 p region between the curve of the reagent and those of the chelates. The chelates, for example, no longer have the broad absorption band a t 2530 kayser since the 0-H bond has been broken in the formation of the metal complex. Here also the chelates exhibit a lower N-H stretching frequency than the reagent itself. 2-(o-Hydroxyphenyl)-imidazolinebehaves very similarly to the corresponding benzimidazole. Although in this instance the N-H stretching vibrations of the chelates are a t a higher frequency than that of the reagent, they are still a t a much lower frequency than one would expect for a "free" N-H stretching vibration. Interpretation of N-H Frequency Shift.-Flett 14 has examined a series of para-substituted anilines and has observed a regular variation in both the symmetric and anti-symmetric X-H stretching frequencies. Those derivatives having a strong electron withdrawing group in the para-position have their N-H vibrations a t a higher frequency than those with a weak electron withdrawing group in the same position. This can be related to the (14) M. St. C. Flett, T r a n s . Faraday SOL.,4 4 , 707 (1948).

STUDY OF METALCHELATES OF IMIDAZOLE DERIVATIVES

Jan. 20, 1956

263

require a change in bond hybridization. Here one is concerned with a pyrrole type nitrogen which has essentially sp2 trigonal bonds, with its lone pair of electrons perpendicular to the plane of the molecule and capable of delocalization. Coordinating H H H the pyridine nitrogen to a metal will increase the \ \ ; / H-3: contribution of structure (4) to the ground state I II of the molecule. This will decrease the N-H stretching frequency because of the formal positive charge on the pyrrole nitrogen. This same effect is present in the case of p-nitroaniline but is overI shadowed by the change in hybridization. N N Coordination of the ammonia molecule or an NH2 group directly to a metal would also place a formal positive charge on the nitrogen and thereby (1) (2) cause the N-H stretching vibration to shift to lower I n contrast to this, Quagliano, Curran, et aZ.,1KJ6 as has been observed.'6s1e have found from an examination of a large number frequencies For the compounds reported on here the exact of compounds containing ammonia molecules or magnitudes of the shifts are not readily discernible NH2 groups that the formation of a nitrogen to because of the tendency of the molecules to assometal bond results in a decrease in the N-H stretch- ciate with one another. ing frequency. Their shifts in the absorption peaks An attempt to correlate the frequency of the attributed to N-H stretching vibrations increase N-H stretching vibration with the relative stawith increasing charge on the metal ion and with bilities of the metal-reagent complexes does not increasing covalent character of the nitrogen to meet with much success, especially when one tries metal bond. to compare the metal series of one reagent with anFor the data presented a t this time, i t is quite other. Structural differences arising from the probable that coordination of these imidazole de- variance in coordination numbers of these metals rivatives with a metal ion also results in a decrease and the packing of the molecules in the solid state of the N-H stretching frequency, which would might possibly explain these minor variations. also be contrary to the observations on the para- Analysis of these compounds in solution would substituted anilines. clarify the latter difficulty, but the relaI n an analogous manner to p-nitroaniline, one probably tive insolubility of the compounds in suitable solcan write the following resonance structures for the vents provides an experimental obstacle. imidazole derivatives. 8 p Region.-Although the spectra of any parH H ticular reagent and its chelates were generally I I +) similar over the range of about 2-15 p, one band pi centering around 1250 kaiser was found to undergo /"\ a regular shift from one metal chelate to another. The most noticeable shift occurred for the chelates \ of 2-(o-hydroxyphenyl)-benzoxazole,and the exact N N(-) positions of this band are as follows

double bond character of the carbon to nitrogen linkage and is a measure of the relative contribution of resonance structure (2) to the ground state of the molecule.

'\

/--J(4)

(3)

At first glance these appear to be comparable to structures (1) and (2) in having a resonance hybrid containing the group \ /NWH. There is a differ-

/

ence, however. Going from structure (1) to (2)involves a change in hybridization of the amine nitrogen ( i e . , going from p to sp2 type bonds). Increasing the amount of s character in the N-H bond would effectively increase the N-H stretching frequency. A strong electron-withdrawing group in the para-position would tend to increase the contribution of structure (2) to the ground state of the molecule. The influence of bond hybridization on stretching frequency has been exemplified for the C-H stretching vibrations of ethane, ethylene and acetylene (spa, sp2 and sp bonding, respe~tively).'~The C-H stretching frequency is the highest for acetylene. T o go from structure (3) to (4), however, does not (15) J. V. Quagliano. G. V. Svatos and Bro. C. Curran, Anal. Chrm., 26, 429 (1954)

(16) Bro. C. Curran, D. N. Sen,S. Mizushima and J. V. Quagliano, THISJOURNAL, 76, 429 (1954). (171 A. p. Walsh, J. Chem. Soc., 398 (1948).

2-(o-Hydroxypheny1)-benzoxazole Chelates.. . . . . . . . . . . . . . . Cu(I1) Ni(I1) Co(I1) Zn(I1) Pb(I1) Mg(I1) Cd(I1) Mn(I1)

1246 kaiser 1261 kaiser 1252 1250 1249 1236 1253 1248 1248

Mellor and Maley18 have pointed out that the stability of complexes of divalent ions follows the order Cu > Ni > P b > Co > Zn > Cd > Mu > Mg irrespective of the nature of the ligands involved. Charles and FreiserIg have established the probable relative stability order of various divalent metal complexes of 2-(o-hydroxyphenyl)-benzoxazoleand found i t to be essentially the same as the MellorMaley series with the exception of an apparent reversal of order in the case of nickel and cobalt and (18) D. P. Mellor and L. E. Maley, Nature, 169, 379 (1947); 161, 436 (1948). (19) R. G. Cbarles and H. Freiser. A n a l . Chim. Acta, 11, 1 (1954).

PAUL J. MCCARTHY AND ARTHUR E. MARTELL

264

with lead placed somewhere between zinc and magnesium. The relative order obtained from a consideration of frequency shifts is Cu

> Mg > Ni > Co > Zn, Cd, M n > P b

The order is practically the same as the usual stability order of divalent metal ions18and also that found by Charles and Freiser.lg The most notable difference is the variance in the position of magnesium. The anomalous magnitude of the frequency shift in the magnesium chelate might be related to the relatively low mass of magnesium. The same type of shift for this peak has also been observed for the chelates of 2-(o-hydroxyphenyl)derivatives of benzimidazole, imidazoline, benzothiazole and benzothiazoline. I n each series the copper chelate, which is the

1701. 7s

most stable of the reagent-metal complexes, has its 1250 kayser absorption peak a t the highest frequency, while the zinc chelate, which is generally less stable than that of the nickel or cobalt, is a t the lowest frequency. This behavior is very similar to that observed for the chelates of 8-hydroxyquinoline and derivatives.20 Acknowledgment.-The authors are indebted to Professor Mary E. Warga of the Physics Department for the use of the infrared spectrometer and to Professor Alfred Berger of the Engineering Department for the use of the press. They gratefully acknowledge the financial assistance of the U. S. Public Health Service. (20) R . G. Charles, H. Freiser. R. Friedel, L. E. Hilliard and W. D. Johnston, Specfrochimica acfa, 8,in press (1956).

PITTSBURGH, PA.

[CONTRIBUTION FROM THE DEPARTMENT O F CHEMISTRY O F CLARK UNIVERSITY]

Dipole Moments of Metal Chelate Compounds. I. Analog of Bisacetylacetone-ethylenediimine BY PAULJ. MCCARTHY AND ARTHUR E. MARTELL RECEIVED JANUARY 31, 1955 Dielectric constants of benzene solutions of bisacetylacetone-ethylenediimine,bisacetylacetonepropylenediirnine,bistrifluoroacetylacetone-ethylenediimine,bisbenzoylacetone-ethylenediimine,bisbenzoylacetonepropylenediimine and bisbenzoylacetonetrimethylenediirnine have been measured, and the values of the corresponding molar polarizations and dipole moments are reported. The magnitudes of the dipole moments are interpreted in terms of rotation of two hydrogenbonded chelate rings about the alkylene bridge.

This is the first of a series of research reports on the dipole moments of chelating agents and metal chelate compounds. The analogs of bisacetylacetone-ethylenediimine were chosen as promising reagents with which to begin this program because the structures involved are relatively simple, and because many of the corresponding metal chelate compounds are soluble in benzene. The tetradentate ligands synthesized for this investigation are summarized in formula I. The names of the comR”

I

pounds, the substituents R, R’ and R”, and the values of n are given in Table I. Of these substances, only the parent compound, bisacetylacetone-ethylenediimine has been previously reported.2 The trifluoro derivative also has been prepared by Martell, Belford and Calvine3 Experimental Apparatus.-A sensitive heterodyne-beat apparatus was set up with a variable-frequency oscillator and a crystal(1) (a) Abstracted from a dissertation submitted by Paul J. McCarthy, t o t h e Faculty of Clark University in partial fulfillment of the requirements for t h e degree of Doctor of Philosophy. (b) This work was supported by a grant from Research Corporation and by t h e Office of Ordnance Research under Contract No. DA19-020-ORD-3243. ( 2 ) A. Combes and C. Combes, C o m p f . r e n d . , 108, 1252 (1889) (3) A. E . Martell, L. Belford and M. Calvin, unpublished results.

TABLE I LIC:ANDSTRUCTURES n

R

1 1

CHa CH, CHI

I

R’

R”

CH? H CHB CHa CFI H

Name of ligand

Bisacetylacetone-ethylenediimine Bisacetylacetonepropylenediimine Bistrifluoroacetylacetone-ethylenediimine Bisbenzoylacetone-ethylenediimine Bisbenzoylacetonepropylenediimine Bisbenzoylacetonetrimethylenediimine

controlled, fixed-frequency oscillator, both of which were fed into a stable receiver with a loudspeaker as detector. Since the beats near the null point were clearly audible, good precision (1part in lo8)was obtained in all capacitance measurements. The fixed-frequency oscillator was constructed according to the circuit of the Blily Model 1C oscillator, with modification which allowed the use of the oscillator at 100 kc. per second as well as the normal operating frequencies of 500 and 1000 kc. per second. The receiver was a National HRO “60” model. The variable-frequency oscillator was a frequency meter, BC-221-T, made by the Zenith Radio Corporation for the Army Signal Corps. A General Radio Type 722-D 1100 ppf. precision condenser was connected to the frequency meter in parallel with the experimental dielectric constant cell. The precision condenser was internally calibrated a t 10 ppf. intervals in a manner similar to that suggested by Smyth.‘ The corrections were plotted against readings of the condenser and were used to correct all subsequent capacitance measurements. The glass dielectric constant cell contained three rigidlymounted platinum cylinders, spaced 0.07 cm. apart. The inner and outer cylinders were connected together and (4) C. P.Smyth, “Determination of Dipole Moments,” in “Physical Methods of Organic Chemistry,” A. Weissberger, Editor, Vol. I. Part 11, Interscience Publishers, New York, N. Y.,1945, pp. 1005 ff.